Most human cancers arise from epithelial cells and become life- threatening when they cross the basement membrane (BM), invade locally and metastasize. An important unresolved problem is what causes carcinoma cells to transgress the BM barrier. Our general hypothesis is that carcinoma cells use migration mechanisms similar to normal epithelial cells, and that we lack understanding of cancer invasion because of incomplete knowledge of the mechanisms controlling normal epithelial cells migration across BM. To test this hypothesis, we propose studies on two epithelial integrins, alpha6beta4 and alpha3beta1, circumstantially associated with invading carcinomas. Both of these integrins bind laminin-5 (Ln-5), an extracellular matrix component of BM, also associated with invasion. Using as a model a non-tumorigenic epithelial cell line, HaCaT, we found that alpha6beta4 stimulates static adhesion and inhibits migration on Ln-5 via phosphatidylinositol 3-kinase (PI3-K). In contrast, alpha3beta1 mediates Ln-5 migration via PI4-K. Our hypothesis is that these integrin-linked signaling pathways are subverted in carcinoma cell invasion, particularly squamous and breast carcinoma. Our Specific Aims are designed to test this hypothesis by dissecting molecular details of these pathways, and then using them to modify invasive behavior of squamous or breast carcinoma cells, ultimately in animal models for metastasis. In Aim 1, we will test, in HaCaT cells, the role of focal adhesion kinase and tetraspanins in linking to PI4-K the alpha3beta1- dependent Ln-5 migration stiumulated by activating integrin antibody TS2/16, and will test the role of differentiation and protein kinase C in mediating alpha3beta1, PI4-K dependent migration induced by serum starvation. In Aim 2, we will test in HaCaT cells whether alpha6beta4 clustering activates PI3-K, inhibiting migration and stimulating static adhesion, and whether alpha6beta4-dependent PI3-K activation requires localization to ventral plasmembrane domains, or cooperation of tyrosine-kinase receptors as Her2/neu, via simultaneous engagement on Ln-5. For these studies, we will use Transwell migration assays, transfection of constitutively active and dominant negative variant of appropriate signaling molecules, as well as receptor and ligand fusion proteins. In Aim 3, we will test the general applicability to cancer invasion of these migratory mechanisms first in a limited panel of carcinoma lines, then on a larger scale by detecting altered expression of their components via DNA microarray hybridization. In Aim 4, we will transfect critical cDNA constructs we will have defined to test whether they increase invasiveness of select cell lines in BM invasion assays, in animal metastasis models and in spontaneously arising carcinomas in transgenic mice.